Perinatal care at the borderlines of viability demands a delicate balance between parents’ wishes and autonomy, biological feasibility, clinicians’ responsibilities and expectations, and the prospects of an acceptable long-term outcome — coupled with a tolerable margin of uncertainty.
A multi-professional workshop with consumer involvement was held in February 2005 to agree on management of this issue in New South Wales and the Australian Capital Territory. Participants discussed and formulated consensus statements after an extensive consultation process.
Consensus was reached that the “grey zone” is between 23 weeks’ and 25 weeks and 6 days’ gestation. While there is an increasing obligation to treat with increasing length of gestation, it is acceptable medical practice not to initiate intensive care during this period if parents so wish, after appropriate counselling.
Poor condition at birth and the presence of serious congenital anomalies have an important influence on any decision not to initiate intensive care within the grey zone.
Women at high risk of imminent delivery within the grey zone should receive appropriate and skilled counselling with the most relevant up-to-date outcome information. Management plans can thus be made before birth. Information should be simple, factual and consistent.
The consensus statements developed will provide a framework to assist parents and clinicians in communication, decision making and managing these challenging situations.
Perinatal care of the extremely low gestation and extremely low birthweight infant has improved. Although the boundary of viability has been challenged and has shifted over the past two decades,1-4 there is substantial risk of long-term neurodevelopmental disabilities in very premature babies who survive.5-9 A number of international guidelines on managing premature babies at the threshold of viability, formulated predominantly by physician expert panels, have been published. These guidelines acknowledge that, within boundaries set by birthweight and gestational age, there is a discretionary zone (a “grey zone”) in which attitudes of parents and clinicians play an important role (Box 1).10-14
The debate on management of neonates at the threshold of viability is an emotive and complex issue that involves families, clinicians and other health care professionals.15-19 In view of this, and in contrast with previously published guidelines prepared by expert panels, it was considered desirable to conduct a consensus workshop to bring together the views of multiple stakeholders, including consumers. The workshop, with participants from New South Wales and the Australian Capital Territory, had two aims:
To produce consensus statements to support clinicians and parents in dealing with the challenging scenarios encountered at the borderlines of viability; and
To agree on accurate, meaningful and consistent information across NSW and the ACT for clinicians, parents and prospective parents of extremely premature infants.
The limits of viability have until recently been addressed in terms of gestational age or birthweight in relation to mortality and long-term outcome. The concept of a gestational “grey zone” was acknowledged by workshop participants as clinically more practical and relevant to perinatal care, as varying biological and clinical situations can render “cut-off” limits impractical and conflict-provoking.
In planning the workshop, the organising group was cognisant of adult learning principles.20 All participants would be actively involved and have experience of the dilemma in practice. It was crucial to achieve meaningful consensus by involving as many well informed stakeholders as possible, with a spread of urban and rural professionals as well as parents. Each of the 10 tertiary neonatal intensive care units in NSW and the ACT were invited to send professional delegates from five disciplinary groups (fetomaternal obstetrics, midwifery, neonatology, neonatal nursing, and developmental teams/other allied health professionals). Representation was also invited from parental groups, professional associations, medical and nursing colleges, and rural or regional clinical practices.
The workshop was held in February 2005 over one and a half days. The program included presentation of outcome data for very premature births along with ethical, parental and medicolegal viewpoints to stimulate discussion and to identify a preliminary grey zone.
It was agreed that the most relevant and up-to-date NSW and ACT survival and outcome data should form the basis for consensus discussions. The analysis included 897 births of premature infants between 22 weeks’ gestation (220) and 25 weeks and 6 days’ gestation (256) to NSW and ACT resident mothers between 1 January 1998 and 31 December 2000.
Follow-up assessments of the surviving children were performed at 2–3 years of age, corrected for prematurity. Most children were assessed at a tertiary centre, but about 10% who lived too far from a major hospital were assessed by their paediatrician or general practitioner. Assessments consisted of a neurological examination and a standardised psychological assessment (the Griffiths Mental Development Scales21 and the Bayley Scales of Infant Development, second edition22).
Categories of functional disability23 were defined as:
None/minimal: developmental quotient above 1 SD below the mean;
Mild: developmental delay (developmental quotient between 1 SD and 2 SD below the mean), or mild cerebral palsy;
Moderate: developmental delay (developmental quotient between 2 SD and 3 SD below the mean), moderate cerebral palsy (able to walk with the assistance of aids), or sensorineural or conductive deafness requiring amplification with bilateral hearing aids or unilateral/bilateral cochlear implant;
Severe: developmental delay (developmental quotient more than 3 SD below the mean), bilateral blindness with visual acuity < 6/60 in the better eye, or severe cerebral palsy (unable to walk with the assistance of aids).
Two interactive large-group sessions — panel discussion and discussion of hypothetical case scenarios — were held to identify key issues pertinent to care around the borderlines of viability. Both sessions were facilitated by an external facilitator.
Hypothetical case scenarios (compiled by the workshop organising committee) highlighted common practical issues and forced participants to confront the way in which decisions have to be made in the clinical area. An electronic multigrade response system (hand-held Digivote [Bramshaw ICS Conference Communications, Melbourne, Australia]) was used by participants to help compile responses. The tracking of each participant’s discipline was based on the delegate’s first answer in identifying his or her discipline or group.
After the large-group discussions, delegates were split into multi-professional groups of 10, each of which was allocated a question or issue developed during the panel session and asked to formulate a response suitable for inclusion in a consensus document. Each group’s consensus statement was presented to the main group for refinement and approval, with the Digivote device used to record each participant’s response on a five-point Likert scale: 1, strongly agree; 2, agree; 3, disagree; 4, strongly disagree; or 5, unable/don’t wish to comment.
Consensus was considered to be unequivocally reached if the combined “agree” and “strongly agree” votes were over 90%, while agreement of 75% was a minimum criterion for consensus, given that statements varied in their degree of gravity.
The proposed consensus statements were grouped according to similarity of themes and content. Statements with higher voting scores formed the principal statement and incorporated any similar statements. The report was circulated for consultation within the committees of the Pregnancy and Newborn Services Network, along with various consumer and professional groups (see Acknowledgements). It was also reviewed by the NSW Health Clinical Ethics Advisory Panel. The process was completed in November 2005.
The workshop was attended by 112 delegates. Participants, according to discipline, were as follows: obstetricians, 10 (9%); midwives, 19 (17%); neonatologists, 28 (25%); neonatal nurses, 29 (26%); allied health professionals or members of a developmental team, 11 (10%); perinatal care administrators (non-clinical), 8 (7%); and parent and community advocates, 7 (6%). Consumers who attended the workshop were offered, and accepted, the opportunity to debrief with a skilled social worker after the workshop.
Statistics on births at 220 to 256 weeks’ gestation in NSW and the ACT are shown in Box 2. Infants born between these gestational ages who were not admitted to a neonatal intensive care unit (NICU) did not survive. With increasing gestational age, an increasing number of infants were born alive and admitted to an NICU.
Of 342 infants admitted to an NICU, 183 survived to hospital discharge. One infant died at 5 months of age, leaving 182 infants available for follow-up, of which 166 (91%) were assessed at 2–3 years (the remainder being lost to follow-up). The average birthweight for children assessed was 750 grams, compared with 713 grams for those lost to follow-up. The average gestational age at birth for both groups was 24.5 weeks.
With increasing gestational age, the proportion of infants diagnosed with mild, moderate or severe functional disability decreased. A comparison was made to a control group of 460 singleton term infants without a major congenital anomaly assessed at 3 years of age using the same assessment methods.24 Extremely premature infants were more likely than term infants to have a mild (13% v 2%), moderate (10% v 1%) or severe (16% v 0.5%) functional disability.
From 10 questions debated by 12 groups, a total of 40 draft statements were proposed, refined and voted on. Acceptance of over 90% was achieved for 26 proposed statements, with 17 receiving acceptance of over 90% across all discipline groups.
Final consensus statements derived from the workshop are shown in Box 3, Box 4 and Box 5. It was established early (voting results not shown) that factors including planned pregnancy, assisted conception, maternal age, maternal illness, antenatal complications and fetal conditions or compromise carried significant weight in the decision-making process. It was acknowledged that unexpected births outside tertiary perinatal centres are even more complex than births in major centres. The local team’s skills, time factors and transport logistical issues have a pivotal impact on the likelihood of a favourable outcome.
Participants were asked the highest gestational age at which they would agree, if requested by parents, not to initiate resuscitation: 100% would agree at 230–6 weeks; 72% at 250–6 weeks; and 36% at 260–6 weeks. The break-even point appears to be between 250 and 266 weeks’ gestation, below which there was a majority trend towards agreeing not to resuscitate (Box 6).
An alternative scenario attempted to remove the influence of parental request (by hypothesising a situation prior to meeting the parents). Over 90% of participants said they would not wish to initiate resuscitation at 230–6 weeks. This fell to 40% at 245 weeks, and only 20% said they would not wish to offer active treatment at 250–6 weeks (Box 7).
Participants were asked to rank the influence of certain factors on making a decision about whether to resuscitate an infant born with a known gestation of 234 weeks. Parental wishes and condition at birth were ranked as the most important issues (Box 3). Other factors included birthweight and absence of congenital abnormality. The availability of resources was ranked the least important factor. Neonatal events that would influence the offering of withdrawal of intensive care included Grade III and Grade IV brain haemorrhage (ie, major intraventricular and/or parenchymal haemorrhage) (ranked highest), ventilator dependence at 36 weeks, and the presence of necrotising enterocolitis requiring surgery.
The incidence of prematurity is currently 6.7% in NSW and the ACT.25 The risk of birth within the gestational grey zone defined in the workshop is 1 in 100, which is higher than the risk for Down syndrome, about which mothers are more likely to receive counselling. The workshop participants overwhelmingly endorsed an effort for general education in the community regarding premature births (Box 4).
The agreed grey zone was identified as between 230 and 256 weeks. The fact that gestational age assessment is not always accurate may confound management decisions. When other serious fetal conditions exist, the management principles recommended here could apply beyond this gestational range. Counselling on the options of active resuscitation and treatment within the grey zone should take into account the infant’s condition at birth, serious anomalies and other relevant factors. It is not appropriate to discuss health care resource implications during counselling. Institutions should develop decision-support management policies and procedures that are culturally sensitive and appropriate for the family and local community (Box 5).
The perinatal care of women and their babies at the borderlines of viability is a complex issue that deserves much ethical debate and consideration.24-29 The risk of infant death or survival with long-term sequelae increases acutely with each weekly (and even daily) reduction in gestational age at birth.7-9 In contrast to guidelines developed by expert panels,10-14 our consensus workshop was conducted as an interactive forum involving well informed members of multidisciplinary perinatal care teams as well as community representatives and parent advocates.
In our workshop, the debate focused on perinatal scenarios and was based on the best antenatal estimation of gestational age rather than birthweight. While it is accepted that female premature infants generally have better outcomes, this factor was not included in discussions, as sex is not usually known before birth. Poor condition at birth correlates with early mortality and morbidity.30,31 Given the uncertainties surrounding very premature births, workshop participants stressed the importance of appropriate counselling in clinical practice and developing management plans before birth.
We took the approach of establishing a grey zone of borderline viability. The grey zones defined in previous guidelines (ranging from 22 to 28 weeks) are summarised in Box 1. Common to all guidelines, parental involvement was emphasised. Each suggested the use of institutional or local outcome data to establish more precise management guidelines within the grey zone. There are inherent pitfalls in generalisation using non-local NICU outcome data, as unique local factors may influence survival and long-term outcome. For example, a US study showed that non-white infants were more likely to survive than white infants, although they did not have a better developmental outcome.7 Furthermore, NICU collaborative networks that receive data contributions from multiple geographically distant centres show differences in outcome between centres.32
Our consensus workshop used regional outcome data that included all births, live births, NICU admissions, survivors and follow-up assessments. Participants’ responses to the case scenarios clearly showed a gradual increase in the acceptance of active intervention with increase in gestational age. Results supported the approach of defining a grey zone rather than a specific cut-off limit. Nonetheless, it was established that the vast majority (> 90%) of workshop participants would not wish to initiate active treatment before 24 weeks, and 100% would agree with parents if non-intervention was requested at this stage.
The hypothetical scenario of “no parental influence” explored the clinician’s own perception of or inclination for intervention. In reality, this scenario does not exist. Nevertheless, clinicians may encounter situations in which there has been limited opportunity for discussion, counselling or preparation of parents.
A complicating factor in the debate about management of extremely premature infants is that clinicians, nurses and parents tend to have differing views about the advisability of active intervention and the health-related quality of life of surviving children. Studies surveying parents of very premature infants and surviving teenage children who had themselves been extremely low birthweight infants (less than 1000 g) have shown that consumers (parents and teenagers) are more likely than health professionals to believe that treatment was worthwhile.33-37 Australian surveys38-41 indicate that neonatologists, neonatal nurses and obstetricians tend to underestimate survival and overestimate likely disabilities. Of the three, neonatologists’ estimates of outcomes are closest to the true picture.
Another issue is the degree to which outcomes are influenced by clinicians’ attitudes. It has been shown in a population-based comparison of two perinatal management strategies that proactive management is associated with better survival of infants born in the grey zone gestational age range.42 The workshop participants considered ongoing audit of outcomes and regular updates of information as essential.
The principle of non-directive counselling (and the difficulty of achieving it) was discussed during the workshop. In a recent survey regarding counselling of parents within the grey zone, substantial differences were noted between clinicians’ and parents’ perceptions.43 Although clinicians aimed to conduct counselling sessions in a non-directive way and believed they had done so, most parents perceived the experience as directive counselling.
Workshop participants agreed that consistent, transparent information should be shared between parents and members of the perinatal team. Information on survival and long-term outcomes of very premature infants was best provided in a written form available to parents. One outcome of the workshop was to update and revise — with substantial consumer input — the NSW and ACT parent information booklet for extremely premature births.44 The objective of the booklet is to provide simple and factual information written in appropriate language for parents. It is envisaged that these efforts will be of great benefit to the counselling process.
Participants also discussed whether all pregnant women should be given information about the risks associated with very premature birth. The incidence of premature birth at the borderlines of viability (1 in 100) is higher than that of aneuploidy. While it appears logical to provide consumer information on premature birth, its presumed benefits need to be evaluated. In contrast to aneuploidy, for which early detection and intervention are possible, there is no effective prevention or treatment for extremely premature birth. Participants suggested that all pregnant women should be given some idea of premature birth rates and information about symptoms of premature labour and other antenatal complications. Early medical attention may then be sought if these symptoms occur.
1 Summary of published guidelines on managing premature infants at the threshold of viability (ie, in the “grey zone”)
2 “Grey zone” births* and outcomes, New South Wales and Australian Capital Territory, 1998–2000
3 Consensus statement: gestational “grey zone”
4 Consensus statement: all pregnancies
5 Consensus statement: clinical approach within the “grey zone”
6 Proportion of participants who would “agree” or “strongly agree” to non-resuscitation if requested by parents